Read more
The work is dedicated to the development of a software and hardware complex to speed up the process of text classification. Within the framework of the work, the basic concepts of the subject area are reviewed. Carrying out detailed descriptions of methods and algorithms of various types that are used in the classification of texts, their advantages and shortcomings are revealed. Based on the analysis conducted, a refined mathematical model is presented using filtering space for category symbols. Within the framework of this work, methods for reducing the cost of hours spent on classifying texts from both software and hardware solutions are presented. Software features include modernized algorithms for advanced text processing and filtering of subject areas. The hardware is based on the use of FPGA power to implement parallel data processing. The peculiarities and results of their vicorization are induced in the robot. Based on the obtained results, a software and hardware complex was developed and research was completed. The results showed that in order to reduce hours spent on classifying texts, it is possible to analyze both the entire complex as a whole, and whether the solutions are presented in the work in a complete combination.
About the author
Tetjana Vasil¿vna Golub, Ph.D., professeure agrégée du Département des systèmes informatiques et des mesures de l'Université nationale « Polytechnique de Zaporizka ».¿rina Jak¿vna Zelen'ova, Ph.D., professeure agrégée, professeure agrégée du Département des systèmes informatiques et des mesures de l'Université nationale.Mar¿ja Jur¿¿vna Tjagunova, Ph.D., professeure agrégée, professeure agrégée du Département des systèmes informatiques et des mesures de l'Université nationale.
